700.27 and 701.18: Local Amendment

[charlie b]

The City of Seattle is in the process of adopting the 2008 NEC, with its own local amendments. One controversial topic under consideration is selective coordination. A proposed amendment would treat selective coordination in such a way as to allow an emergency (or legally required standby) system to include breakers that do not coordinate in their instantaneous trip regions. What the proposed local code amendment would allow is that, under engineering supervision, the selective coordination requirements of 700.27 and 701.18 would be treated as being satisfied, without requiring consideration of a breaker’s instantaneous trip region (i.e., faults with a duration, before OCPD actuation, of 0.1 seconds or less would be permitted to be disregarded)?

Supporters of this amendment appear to be focusing on the low probability that a branch circuit will experience a fault of such high magnitude as to cause the trip of a breaker upstream. Opponents of this amendment appear to be focusing on the potentially severe consequences of a total blackout (i.e., loss of all power in a hospital while surgery is in progress or while a person is on a life-support system).

Anyone have any thoughts or experiences to share?

 

[wirenut1980]

I voted no, because if I were personally the designer, I would not feel comfortable with this exception even if it were legal. I suppose if the maximum fault current available at a point downstream of a branch OCPD were calculated to not trip the instantaneous on an upstream OCPD it would make me feel better, but not good enough to put my stamp on it.

 

[jim dungar]

The State of Wisconsin, voted down an a recommended exception like this at the last possible minute when they adopted the 2008NEC. One rational is that by not having an exact exception the AHJ is now able to accept some amount of risk analysis when deciding what is "selective coordination".

This allows the AHJ to consider the most likely fault current flowing in the circuit and not just the most fault current available at the line side of the breaker. For example: it is very probable that a lighting or receptacle circuit will experience a bolted fault, however the most likely fault to occur on a large feeder circuit is an arcing fault except during periods of re-energizing after planned maintenance when an outage can be tolerated.

 

[brother]

The City of Seattle is in the process of adopting the 2008 NEC, with its own local amendments. One controversial topic under consideration is selective coordination. A proposed amendment would treat selective coordination in such a way as to allow an emergency (or legally required standby) system to include breakers that do not coordinate in their instantaneous trip regions. What the proposed local code amendment would allow is that, under engineering supervision, the selective coordination requirements of 700.27 and 701.18 would be treated as being satisfied, without requiring consideration of a breaker?s instantaneous trip region (i.e., faults with a duration, before OCPD actuation, of 0.1 seconds or less would be permitted to be disregarded)?

Supporters of this amendment appear to be focusing on the low probability that a branch circuit will experience a fault of such high magnitude as to cause the trip of a breaker upstream. Opponents of this amendment appear to be focusing on the potentially severe consequences of a total blackout (i.e., loss of all power in a hospital while surgery is in progress or while a person is on a life-support system).

Anyone have any thoughts or experiences to share?

I would be totally AGAINST THIS, I have seen more than once where a 'branch circuit' cause a total BLACK OUT in a part of a hospital, because the cordination was not done right. This needs to be REQUIRED. I would contact any authorities concerning this. I cant believe they are even considering this!!

 

[RB1]

I am interested in how other jurisdictions are enforcing this requirement as well. I also wonder if the varous Code Making Panels are familiar with Section 90.3(C)-"This Code is not intended as a design specification"

 

[charlie b]

I have heard that the State of Oregon has approved an amendment similar to this. Can anyone confirm or refute this rumor?

 

[charlie b]

I have seen more than once where a 'branch circuit' cause a total BLACK OUT in a part of a hospital, because the coordination was not done right.

Please clarify two aspect of this incident for me.

First, how did the single branch circuit take out both the utility and the backup generator?

Secondly, what do you intend the phrase ?not done right? to mean? Please note that the notion of ?not done right? is different than ?defining right in terms of minimum fault duration.? If the engineer of record made proper design decisions and specified proper breaker settings, and if the installer did not make the correct settings, then I would describe that as ?not done right.? Similarly, if the engineer of record made an error in the calculations, and if the installer did exactly what the engineer specified, I would also describe that as ?not done right,? and cite the engineer as being at fault.

However, as is its normal way, the NEC does not describe what is ?right,? when it says we must selectively coordinate certain overcurrent devices. It says ?do it,? without saying how, and without saying what constitutes adequate coordination. If we require coordination at all possible fault levels, even though the extreme higher fault levels are extremely unlikely to occur, then are we giving our clients a reasonable balance between risks and costs?

There will certainly be some measureable degree of risk to the lives and health of hospital patients, if a fault within a 20 amp branch circuit were to have the potential of tripping the main service breaker. But I submit that there would be a greater risk to the health and safety of the public, if any concerned citizen were to drive a car from their home to the city?s offices, to attend a public meeting during which such an amendment were to be discussed. Car accidents kill and maim a lot of people every day. Electrical faults that blackout hospitals and that result from inadequately designed selective coordination (and from no other cause) do not.

 

[jim dungar]

However, as is its normal way, the NEC does not describe what is ?right,? when it says we must selectively coordinate certain overcurrent devices. It says ?do it,? without saying how, and without saying what constitutes adequate coordination. If we require coordination at all possible fault levels, even though the extreme higher fault levels are extremely unlikely to occur, then are we giving our clients a reasonable balance between risks and costs?.

This is why the AHJ in Wisconsin lobbied against a defined exception which would tie their hands. By leaving the requirements "open" they felt they can administer the Code in the best interests of the public. Another example: when a new emergency distribution system is being added to an existing facility, Wisconsin feels an open requirement maybe able to accommodate the grey areas of selectivity that almost always occur between new and existing equipment.

 

[don_resqcapt19]

It is my opinion that if you really need selective coordination, then it should go to time zero. In many cases, it is really only a high fault condition that causes the upstream device to open...the very condition that would require coordination to time zero as a .1 second time would permit more the upstream device to open.
This is also a breaker manufacturer v fuse manufacturer issue, with the fuse people (the ones who got this rule into the code) in favor of coordination to time zero and the breaker people wanting a .1 second time frame.

 

[charlie b]

Let me split a hair here. I consider the following two things to be very different.

Thing One:
Suppose that when looking at my choice of breakers, I discover that two do not coordinate with each other in the instantaneous range. No surprise there; they seldom do. Specifically, suppose that I discovered that a branch circuit breaker and an upstream feeder to a distribution panel could both open, if a fault in the branch circuit reached a level of 12,000 amps. Suppose further that the results of my fault current calculation showed that the amount of fault current available at that branch circuit was only 9,000 amps. I would declare that this situation is in compliance with the NEC, and that selective coordination has been successfully demonstrated. My reasoning is that I do not have to take into account a fault that cannot possibly occur. The current would never get high enough to cause both breakers to attempt to open at the same time; the downstream breaker would always trip first.

Thing Two:
Same setup, except that the results of my fault current calculation showed that the amount of fault current available at that branch circuit was 13,000 amps. If the proposed Seattle amendment passes, I could declare that this situation is in compliance with the NEC, and that selective coordination has been successfully demonstrated. My reasoning would be that I do not have to take into account a fault that is so unlikely to take place that City has declared the risk to be acceptable.

So what I perceive the City to be contemplating is establishing a degree of acceptable risk.

 

[steve66]

By requiring total coordination, you are ignoring many other aspects of electrical design, including arc flash hazzards, and possible damage to equipment and feeders.

It also ignores the fact that we are frequently adding onto existing systems, and adding into existing buildings. Many times, you just can't achieve total coordination without replacing the entire system - which is usually impossible or impractacle.

I have seem several places where other safety aspects would have to be comprimised to achieve total coordination. Is that what we really want?

Steve

 

[don_resqcapt19]

By requiring total coordination, you are ignoring many other aspects of electrical design, including arc flash hazzards, and possible damage to equipment and feeders.
Steve

That sounds exactly like what a breaker guy said at a presentation I attended, but then the fuse guy got up and said, "gee, we don't have any problems like that with our products".
I don't have enough background to know which side, if either, is correct on this issue. It seems like a lot of the information and rhetoric on the issue is really a turf war between competing products....remember that the selective coordination rule was proposed by a guy who works for Bussmann.

 

[jim dungar]

That sounds exactly like what a breaker guy said at a presentation I attended, but then the fuse guy got up and said, "gee, we don't have any problems like that with our products".
I don't have enough background to know which side, if either, is correct on this issue. It seems like a lot of the information and rhetoric on the issue is really a turf war between competing products....remember that the selective coordination rule was proposed by a guy who works for Bussmann.

I regularly (I finished my latest one last Monday) figure out how to selectively coordinate breaker systems. So I am not a believer that it is still strictly a turf war.

Areas of selective coordination that are not often addressed include:
A) Fuse selectivity ratios are only guaranteed by each individual manufacturer. If the customer changes a fuse from Manufacturer A to one from Manufacturer B the system may longer be coordinated. It seems that the most likely fuses to blow, and therefore need coordination, are the ones most likely to be purchased based on cost not manufacturer.
B) The most probable fault in any circuit is a ground fault. The NEC selective coordination requirement does not apply to ground fault schemes. But there is a minimum .1sec delay between the required multiple levels of ground fault in hospitals. If .1 sec is okay for the most probable faults why not for the most improbable ones?
C) How do you selectively coordinate (4) 800 paralleled generators feeding a switchboard with (1) 300A circuit, (2) 800A circuits, and (1) 1200A?
D) If new service entrance equipment is added to a facility is selective coordination needed with the existing equipment?

 

[don_resqcapt19]

I regularly (I finished my latest one last Monday) figure out how to selectively coordinate breaker systems. So I am not a believer that it is still strictly a turf war.

Most of the information is about this issue is coming from the fuse and breaker people...it sure looks like a turf war to me.

Areas of selective coordination that are not often addressed include:
A) Fuse selectivity ratios are only guaranteed by each individual manufacturer. If the customer changes a fuse from Manufacturer A to one from Manufacturer B the system may longer be coordinated. It seems that the most likely fuses to blow, and therefore need coordination, are the ones most likely to be purchased based on cost not manufacturer.

Yes fuses must be replaced with the same brand and type of fuse.

B) The most probable fault in any circuit is a ground fault. The NEC selective coordination requirement does not apply to ground fault schemes. But there is a minimum .1sec delay between the required multiple levels of ground fault in hospitals. If .1 sec is okay for the most probable faults why not for the most improbable ones?

I don't understand why 700.27 and 701.18 don't apply to ground faults. The wording requires that "overcurrent" devices to be selectively coordinated.
The Article 100 definition of overcurrent includes ground faults.

C) How do you selectively coordinate (4) 800 paralleled generators feeding a switchboard with (1) 300A circuit, (2) 800A circuits, and (1) 1200A?

I guess, as the design engineer, that is your problem.:grin: If you can't comply with 700.27 or 701.18, you would have to redesign your distribution system so you can comply.

D) If new service entrance equipment is added to a facility is selective coordination needed with the existing equipment?

It appears that you do have to coordinate with the existing equipment as the rule says:
"Emergency system(s) overcurrent devices shall be selectively coordinated with all supply side overcurrent protective devices"

 

[jim dungar]

Don,

You missed my intention. I wanted to point out issues that are commonly "forgotten about" but are impacted by the issue of selective coordination and sometimes are reasons for a debate over any possible exception.

 

[don_resqcapt19]

Don,

You missed my intention. I wanted to point out issues that are commonly "forgotten about" but are impacted by the issue of selective coordination and sometimes are reasons for a debate over any possible exception.

Sorry, Jim, I did miss your point.

Most of the "debate" I have seen is not down to the details that you point out. Most of it has been from reps of the two sides saying why it can or cannot be accomplished. I am not sure there is a real safety issue behind the code rules that were put into the code by Bussmann...maybe that is one reason I think it is mostly a turf war.

 

[jim dungar]

I am not sure there is a real safety issue behind the code rules that were put into the code by Bussmann...maybe that is one reason I think it is mostly a turf war.

If you listen to salespeople you will think it is a turf war. But since when have sales people lived in the actual world.:smile:

The requirement is in the NEC. It will probably never be removed from the NEC. Hopefully there will be more exceptions and clarifications. But in the mean time it is not extremely difficult to achieve once the manufacturer of the equipment is selected and the correct equipment can be chosen, since most breaker manufacturers now have actual tested combinations of devices. After all citical power centers have been doing some of it for years

The biggest problem I face is the people that do not even consider coordination in their designs. It is not easy to coordinate 7 protective devices in series regardless if you are using fuses or breakers.

 

[charlie b]

Many thanks for the comments, and the votes, so far.

I do not know how to get access to, or to search, the ROP/ROC information from previous code cycles. Do you need to be an NFPA member? If this is not a difficult or time consuming task, can someone help me find (1) The substantiation(s) that got this into the code in the first place, and (2) The substantiation(s) that kept it from being revised to essentially waive the first 0.1 seconds? The Bussman representative who spoke at the recent board meeting in Seattle stated that a code change proposal of this nature has twice failed. I wonder if the reason it failed was that the CMP had a technical objection to the change, or that the proposer did not provide a sufficient substantiation.

By the way, and perhaps I should have mentioned this before, my intent for posting this thread is to present the results to Seattle's "Construction Code Advisory Board," to assist them in making their decision on the matter. I don't plan to give any names.

For the record, IMHO this is not a "legal dispute," in the sense that is restricted by forum rules. I see it rather as being a proposal being considered by a governmental body that is involved in the code making process. So if anyone objects to my using any comments posted herein, I'll not do so.

 

[brother]

Please clarify two aspect of this incident for me.

First, how did the single branch circuit take out both the utility and the backup generator? ]

On one incident I knew of, this was a 'large' piece of equipment. There was no 'generator' this hospital uses another 'service' per se for the emerency power. This equipment was on a 'ats' and the 'normal' trip first on a ground fault, then it switch to e power, and that breaker tripped on ground fault. Both of these were the main feeder breakers and fuss and down stream breakers never did trip. This should not have happen, and besides according to code for a HOSPITAL setting, there should not be any gfpe on the e power side.

Secondly, what do you intend the phrase ?not done right? to mean? Please note that the notion of ?not done right? is different than ?defining right in terms of minimum fault duration.? If the engineer of record made proper design decisions and specified proper breaker settings, and if the installer did not make the correct settings, then I would describe that as ?not done right.? Similarly, if the engineer of record made an error in the calculations, and if the installer did exactly what the engineer specified, I would also describe that as ?not done right,? and cite the engineer as being at fault. ]]

Im not sure who/ or what exactly was not done right. I just know it was not done right. the Main feeder breaker should not trip before the branch breaker when the fault is on the branch side. But I could not even imagine this not being required (seletivity) assuming in other(Non hospital settings) applications for e power for this would reduce the reliability of e power if one decides to have gfpe on e power at all. Last I recall it was OPTIONAL for e power (other non hospital settings). Hopes this answered your questions.

 

[don_resqcapt19]

...
The requirement is in the NEC. It will probably never be removed from the NEC. Hopefully there will be more exceptions and clarifications. ...

Jim,
My point is that the requirement was put into the code by Bussmann without any real substantiation and I am not sure that it really is a safety issue. There was no documentation of real world issues on emergency systems without selective coordination...just a lot of "what ifs". There have been many proposals over the years that had much better substantiations that were rejected by code making panels with the statement that the substantiation was inadequate. It is my opinion that the only real purpose for this proposal was to increase market share for the submitter.
That being said, I agree that it is unlikely to be removed, however it is also my opinion that if you can show that selective coordination is a real safety issue(not sure that can be done), then it has to apply in all cases and down to time zero.
As far as clarification we should expect some in the 2011 code as there are 10 proposals submitted for 700.27. Some are in direct conflict so CMP 13 will have to make some statements that will clarify the issues. In particular there is one to permit the coordination to start at 0.1 seconds and one to make it clear that coordination is required to time zero. There is also one that states in its substantiation that the manufacturers of one type of OCPDs are using the existing section improperly to increase market share.
There is an interesting one from a breaker manufacture that would prohibit the use of "maintenance" switches that are offered by some breaker manufactures to decrease the instantaneous trip so as to reduce the arc flash hazard for live work.